total quality management: focus on six sigma - McGraw

TOTAL QUALITY MANAGEMENT: FOCUS ON SIX SIGMA
chapter 7
chapter
TOTAL QUALITY MANAGEMENT:
FOCUS ON SIX SIGMA
2 74
Total Quality Management
Total quality management defined
Malcolm Baldrige National Quality Award defined
276
Quality Specification and Quality Costs
Developing quality specifications
Cost of quality
Functions of the QC department
279
Six-Sigma Quality
Six-Sigma methodology
Analytical tools for Six Sigma and
continuous improvement
Six-Sigma roles and responsibilities
286
Design quality defined
Conformance quality defined
Quality at the source defined
Dimensions of quality defined
Cost of quality defined
Six Sigma defined
DPMO defined
DMAIC defined
PCDA cycle defined
Continuous improvement defined
Kaizen defined
Black belts, master black belts, and green belts defined
The Shingo System: Fail-Safe Design
Fail-safe procedures defined
Poka-yoke defined
286
ISO 9000
The ISO 9000 series
ISO 9000 certification
289
ISO 9000 defined
ISO 14000 defined
External Benchmarking for Quality Improvement
External benchmarking defined
290
Service Quality Measurement: SERVQUAL
SERVQUAL defined
e-SERVICE QUALITY defined
291
Conclusion
294
Case: Hank Kolb, Director of Quality Assurance
295
Case: Shortening Customers’ Telephone Waiting Time
298
Case: “Hey, Is Anybody There?” An Example of DMAIC at
American Express
7
● ● ● “This big myth is that Six Sigma is about quality control and statistics. It is
that—but it’s much more. Ultimately, it drives leadership to be better by providing tools to think through tough issues. At Six Sigma’s core is an idea that
can turn a company inside out, focusing the organization outward on the
customer.”
—Jack Welch, Straight from the Gut (New York: Warner Business Books, 2001), p. 330.
What is GE doing with Six Sigma under Welch’s successor Jeffery R. Immelt?
More than ever, GE is spending $600 million on Six Sigma projects in 2002—
mostly for the salaries of 4,000 full time experts and 100,000 employees
who have undergone basic training. They not only have targeted finding
$2.5 billion in savings in GE, but are sending out its Sigma Squads to customers such as Dell Computers and Wal-Mart to help them root out what
they estimate to be $1 billion in inefficiencies and waste.
—Michael Arndt, “Quality Isn’t Just for Widgets,” Business Week, July 22, 2002, p. 72.
The opening quote from the former chairman and CEO of General Electric
captures the essence of what made GE’s program so successful, and the
excerpt from a recent Business Week shows that Six Sigma is still a major
part of GE’s operations, and an important part of the American quality
movement. –>
274
section 2
PRODUCT DESIGN AND PROCESS SELECTION
In this chapter, we first review the general subject of total quality management and the quality movement. We then develop the basic features and concepts of the Six Sigma approach
to TQM. We then describe the Shingo system, which takes a unique approach to quality by
focusing on preventing mistakes. This is followed by a review of ISO 9000 standards for
quality certification used by many companies throughout the world. We then provide the
major steps of external benchmarking for quality improvement. We conclude with a presentation of SERVQUAL, a tool designed expressly for measuring quality in service delivery.
T O TA L Q U A L I T Y M A N A G E M E N T
Total quality management
● ● ● Total quality management may be defined as “managing the entire organization so that it excels on all dimensions of products and services that are important to the
customer.” It has two fundamental operational goals, namely
1 Careful design of the product or service.
2 Ensuring that the organization’s systems can consistently produce the design.
These two goals can only be achieved if the entire organization is oriented toward them—
hence the term total quality management. TQM became a national concern in the United
States in the 1980s primarily as a response to Japanese quality superiority in manufacturing
automobiles and other durable goods such as room air conditioners. A widely cited study of
Japanese and U.S. air-conditioning manufacturers showed that the best quality American
products had higher average defect rates than those of the poorest Japanese manufacturers.1
So severe was the quality shortfall in the United States that improving it throughout industry
B
breakthrough
R E A K T H R O U G H
BALDRIGE QUALITY AWARD
The Baldrige Quality Award is given to organizations that have
demonstrated outstanding quality in their products and processes. Three awards may be given annually in each of these
categories: manufacturing, service, small business, and, starting
in 1999, education and health care.
Candidates for the award must submit an application of up
to 75 pages that details the approach, deployment, and results of their quality activities under seven major categories:
Leadership, Strategic Planning, Customer and Market Focus,
Information and Analysis, Human Resource Focus, Process
Management, and Business Results. These applications are
scored on total points out of 1,000 by examiners and judges.
Those who score above roughly 650 are selected for site
visits. Winners selected from this group are then honored at
an annual meeting in Washington, DC. A major benefit to all
applicants is feedback from the examiners, which is essentially an audit of their practices. Many states have used the
Baldrige Criteria as the basis of their own quality award programs. A report, Building on Baldrige: American Quality for
the 21st Century, by the private Council on Competitiveness,
said, “More than any other program, the Baldrige Quality
Award is responsible for making quality a national priority
and disseminating best practices across the United States.”
TOTAL QUALITY MANAGEMENT: FOCUS ON SIX SIGMA
became a national priority, with the Department of Commerce establishing the Malcolm
Baldrige National Quality Award in 1987 to help companies review and structure their
quality programs. Also gaining major attention at this time was the requirement that suppliers demonstrate that they are measuring and documenting their quality practices according
to specified criteria, called ISO standards, if they wished to compete for international contracts. We will have more to say about this later.
The philosophical leaders of the quality movement, notably Philip Crosby, W. Edwards
Deming, and Joseph M. Juran—the so called Quality Gurus—had slightly different definitions of what quality is and how to achieve it (see Exhibit 7.1), but they all had the same
general message: To achieve outstanding quality requires quality leadership from senior
management, a customer focus, total involvement of the workforce, and continuous improvement based upon rigorous analysis of processes. Later in the chapter, we will discuss how
these precepts are applied in the latest approach to TQM—Six Sigma. We will now turn to
some fundamental concepts that underlie any quality effort: quality specifications and quality
costs.
chapter 7
Malcolm Baldrige National
Quality Award
EXHIBIT 7.1
The Quality Gurus Compared
CROSBY
DEMING
JURAN
Definition of quality
Conformance to requirements
A predictable degree of
uniformity and dependability
at low cost and suited to the
market
Fitness for use (satisfies
customer’s needs)
Degree of senior management
responsibility
Responsible for quality
Responsible for 94% of
quality problems
Less than 20% of quality
problems are due to workers
Performance standard/
motivation
Zero defects
Quality has many “scales”;
use statistics to measure
performance in all areas;
critical of zero defects
Avoid campaigns to do
perfect work
General approach
Prevention, not inspection
Reduce variability by
continuous improvement;
cease mass inspection
General management
approach to quality,
especially human elements
Structure
14 steps to quality
improvement
14 points for management
10 steps to quality
improvement
Statistical process control
(SPC)
Rejects statistically acceptable
levels of quality [wants 100%
perfect quality]
Statistical methods of quality
control must be used
Recommends SPC but warns
that it can lead to
tool-driven approach
Improvement basis
A process, not a program;
improvement goals
Continuous to reduce
variation; eliminate goals
without methods
Project-by-project team
approach; set goals
Teamwork
Quality improvement teams;
quality councils
Employee participation in
decision making; break down
barriers between departments
Team and quality circle
approach
Costs of quality
Cost of nonconformance;
quality is free
No optimum; continuous
improvement
Quality is not free; there is
not an optimum
Purchasing and goods
received
State requirements; supplier
is extension of business; most
faults due to purchasers
themselves
Inspection too late; sampling
allows defects to enter system;
statistical evidence and control
charts required
Problems are complex; carry
out formal surveys
Vendor rating
Yes; quality audits useless
No, critical of most systems
Yes, but help supplier
improve
275
276
section 2
PRODUCT DESIGN AND PROCESS SELECTION
Q U A L I T Y S P E C I F I C AT I O N A N D
QUALITY COSTS
● ● ● Fundamental to any quality program is the determination of quality specifications and the costs of achieving (or not achieving) those specifications.
D E V E LO P I N G Q U A L I T Y S P E C I F I C AT I O N S
Design quality
Conformance quality
Quality at the source
Serv
i
ce
Dimensions of quality
EXHIBIT 7.2
The Dimensions of
Design Quality
The quality specifications of a product or service derive from decisions and actions made
relative to the quality of its design and the quality of its conformance to that design. Design
quality refers to the inherent value of the product in the marketplace and is thus a strategic
decision for the firm. The common dimensions of design quality are listed in Exhibit 7.2.
Conformance quality refers to the degree to which the product or service design specifications are met. The activities involved in achieving conformance are of a tactical,
day-to-day nature. It should be evident that a product or service can have high design quality
but low conformance quality, and vice versa.
Quality at the source is frequently discussed in the context of conformance quality.
This means that the person who does the work takes responsibility for making sure that his
or her output meets specifications.
Where a product is involved, achieving the quality specifications is typically the responsibility of manufacturing management; in a service industry, it is usually the responsibility of
the branch operations management. Exhibit 7.3 shows two examples of the dimensions of
quality. One is a stereo amplifier that meets the signal-to-noise ratio standard; the second is
a checking account transaction in a bank.
DIMENSION
MEANING
Performance
Primary product or service characteristics
Features
Added touches, bells and whistles, secondary characteristics
Reliability
Consistency of performance over time, probability of failing
Durability
Useful life
Serviceability
Ease of repair
Response
Characteristics of the human-to-human interface (speed, courtesy, competence)
Aesthetics
Sensory characteristics (sound, feel, look, and so on)
Reputation
Past performance and other intangibles (perceived quality)
SOURCE FOR EXHIBITS 7.1 AND 7.2: MODIFIED FROM JOHN S. OAKLAND, TOTAL QUALITY MANAGEMENT (LONDON: HEINEMANN PROFESSIONAL
PUBLISHING LTD., 1989), PP. 291–92.
EXHIBIT 7.3
Examples of Dimensions
of Quality
MEASURES
DIMENSION
PRODUCT EXAMPLE:
STEREO AMPLIFIER
SERVICE EXAMPLE:
CHECKING ACCOUNT AT A BANK
Performance
Signal-to-noise ratio, power
Time to process customer requests
Features
Remote control
Automatic bill paying
Reliability
Mean time to failure
Variability of time to process requests
Durability
Useful life (with repair)
Keeping pace with industry trends
Serviceability
Modular design
Online reports
Response
Courtesy of dealer
Courtesy of teller
Aesthetics
Oak-finished cabinet
Appearance of bank lobby
Reputation
Market leader for 20 years
Endorsed by community leaders
SOURCE: MODIFIED FROM PAUL E. PISEK, “DEFINING QUALITY AT THE MARKETING/DEVELOPMENT INTERFACE,” QUALITY PROGRESS, JUNE 1987, PP. 28–36.
TOTAL QUALITY MANAGEMENT: FOCUS ON SIX SIGMA
chapter 7
277
GLOBAL MANUFACTURING FOR MENNEN SPEEDSTICK® DEODORANT AT A SINGLE LOCATION
IN MORRISTOWN, NJ, SERVES 48 DIFFERENT
COUNTRIES. THE TECHNICIAN IS INSPECTING
THE MULTILANGUAGE LABELING THAT ENABLES
THE SAME PRODUCT TO BE SOLD IN VARIOUS
Glob
COUNTRIES.
al
Both quality of design and quality of conformance should provide products that meet
the customer’s objectives for those products. This is often termed the product’s fitness for
use, and it entails identifying the dimensions of the product (or service) that the customer
wants (that is, the voice of the customer) and developing a quality control program to ensure
that these dimensions are met.
COST
OF
QUALITY
Although few can quarrel with the notion of prevention, management often needs hard numbers to determine how much prevention activities will cost. This issue was recognized by
Joseph Juran, who wrote about it in 1951 in his Quality Control Handbook. Today, cost of
quality (COQ) analyses are common in industry and constitute one of the primary functions
of QC departments.
There are a number of definitions and interpretations of the term cost of quality. From
the purist’s point of view, it means all of the costs attributable to the production of quality
that is not 100 percent perfect. A less stringent definition considers only those costs that
are the difference between what can be expected from excellent performance and the current costs that exist.
How significant is the cost of quality? It has been estimated at between 15 and 20 percent
of every sales dollar—the cost of reworking, scrapping, repeated service, inspections, tests,
warranties, and other quality-related items. Philip Crosby states that the correct cost for a
well-run quality management program should be under 2.5 percent.2
Three basic assumptions justify an analysis of the costs of quality: (1) failures are caused,
(2) prevention is cheaper, and (3) performance can be measured.
The costs of quality are generally classified into four types:
1
2
3
4
Appraisal costs. Costs of the inspection, testing, and other tasks to ensure that the
product or process is acceptable.
Prevention costs. The sum of all the costs to prevent defects, such as the costs
to identify the cause of the defect, to implement corrective action to eliminate the
cause, to train personnel, to redesign the product or system, and to purchase new
equipment or make modifications.
Internal failure costs. Costs for defects incurred within the system: scrap, rework,
repair.
External failure costs. Costs for defects that pass through the system: customer
warranty replacements, loss of customers or goodwill, handling complaints, and
product repair.
Cost of quality
278
section 2
PRODUCT DESIGN AND PROCESS SELECTION
EXHIBIT 7.4
CURRENT MONTH’S
COST
PERCENTAGE
OF TOTAL
Prevention costs
Quality training
Reliability consulting
Pilot production runs
Systems development
$ 2,000
10,000
5,000
8,000
1.3%
6.5
3.3
5.2
Total prevention
25,000
16.3
Appraisal costs
Materials inspection
Supplies inspection
Reliability testing
Laboratory testing
6,000
3,000
5,000
25,000
3.9
2.0
3.3
16.3
Total appraisal
39,000
25.5
15,000
18,000
12,000
6,000
9.8
11.8
7.8
3.9
51,000
33.3
14,000
6,000
3,000
10,000
5,000
9.2
3.9
2.0
6.5
3.3
38,000
24.9
$153,000
100.0
Quality Cost Report
Internal failure costs
Scrap
Repair
Rework
Downtime
Total internal failure
External failure costs
Warranty costs
Out-of-warranty repairs and replacement
Customer complaints
Product liability
Transportation losses
Total external failure
Total quality costs
Serv
i
ce
Exhibit 7.4 illustrates the type of report that might be submitted to show the various costs
by categories. Prevention is the most important influence. A rule of thumb says that for
every dollar you spend in prevention, you can save $10 in failure and appraisal costs.
Often increases in productivity occur as a by-product of efforts to reduce the cost of
quality. A bank, for example, set out to improve quality and reduce the cost of quality and
found that it had also boosted productivity. The bank developed this productivity measure
for the loan processing area: the number of tickets processed divided by the resources required (labor cost, computer time, ticket forms). Before the quality improvement program,
the productivity index was 0.2660 [2,080/($11.23 × 640 hours + $0.05 × 2,600 forms +
$500 for systems costs)]. After the quality improvement project was completed, labor time
fell to 546 hours and the number of forms fell to 2,100, for a change in the index to 0.3088,
an increase in productivity of 16 percent.
FUNCTIONS
OF THE
Q C D E PA RT M E N T
Although the focus of this chapter is on corporatewide quality programs, it is useful to comment on the functions of QC departments.
The typical manufacturing QC department has a variety of functions to perform. These
include testing designs for their reliability in the lab and the field; gathering performance data
on products in the field and resolving quality problems in the field; planning and budgeting
the QC program in the plant; and, finally, designing and overseeing quality control systems
and inspection procedures, and actually carrying out inspection activities requiring special
technical knowledge to accomplish. The tools of the QC department fall under the heading
of statistical quality control (SQC) and consist of two main sections: acceptance sampling
and process control. These topics are covered in the technical note to this chapter.
TOTAL QUALITY MANAGEMENT: FOCUS ON SIX SIGMA
chapter 7
THE THEORY BEHIND SIX SIGMA
Six Sigma reflects the goal of having the customer’s specification limits of the items produced by a process be twice the
natural variation (±3σ) of the process outputs (or, to put it
another way, the process variation to be half the specification
limits). For example, suppose the outputs of a process are normally distributed as shown above line A in Exhibit 7.5. By the
definition of ±3σ (σ is the standard deviation), 99.73 percent of
the outputs are within ±3σ of the mean and, therefore,
.27 percent are outside ±3σ of the mean. Thus, if the specification limits of the product are set equal to ±3σ for the process, we could expect .27 percent of the outputs to be out of
specification. That is, we could expect 2.7 units per thousand
or 2,700 units or parts per million (ppm) to be out of specification. Now suppose that we decide that a defect rate of
2,700 ppm is too high. If the specification limits are kept at the
same place and the process is improved so the output variation is much less, the probability of producing a unit out of
specification will go down. This is shown below line B in the
exhibit. In particular, suppose the process is improved to
the point where the interval of natural variation (±3σ) of the
process is half of the interval of the specification limits (which
then, by definition, will be ±6σ for the process outputs). Then,
the probability of producing a unit outside the ±3σ interval
remains .0027 (by definition of ±3σ), but the probability of
having a part produced out of the specification interval is an
order of magnitude less—about two parts per billion (by the
definition of ±6σ). Details about the statistical characteristics
of Six Sigma are covered in Technical Note 7.
Comparison of Three Sigma and Six Sigma Distributions
EXHIBIT 7.5
2700 ppm outside 3␴ limits
A
⫺3␴
⫹3␴
spec limits
B
⫺6␴
2700 ppm outside
spec limits
⫹6␴
2700 ppm outside 3␴ limits
2 ppb outside 6␴ limits
2 ppb outside spec limits
⫺3␴
⫹3␴
S I X- S I G M A Q U A L I T Y
● ● ● Six-Sigma refers to the philosophy and methods companies such as General
Electric and Motorola use to eliminate defects in their products and processes. A defect is
simply any component that does not fall within the customer’s specification limits. Each
step or activity in a company represents an opportunity for defects to occur and Six-Sigma
programs seek to reduce the variation in the processes that lead to these defects. Indeed,
Six-Sigma advocates see variation as the enemy of quality, and much of the theory underlying Six Sigma is devoted to dealing with this problem. A process that is in Six-Sigma
control will produce no more than two defects out of every billion units.
Six Sigma
279
280
section 2
DPMO
PRODUCT DESIGN AND PROCESS SELECTION
One of the benefits of Six-Sigma thinking is that it allows managers to readily describe
the performance of a process in terms of its variability and to compare different processes
using a common metric. This metric is defects per million opportunities (DPMO). This
calculation requires three pieces of data:
1 Unit. The item produced or being serviced.
2 Defect. Any item or event that does not meet the customer’s requirements.
3 Opportunity. A chance for a defect to occur.
A straightforward calculation is made using the following formula:
DPMO =
Number of defects
× 1,000,000
Number of opportunities for error per unit × Number of units
EXAMPLE 7.1
The customers of a mortgage bank expect to have their mortgage applications processed within 10 days
of filing. This would be called a critical customer requirement, or CCR, in Six-Sigma terms. Suppose all
defects are counted (loans in a monthly sample taking more than 10 days to process), and it is determined
that there are 150 loans in the 1,000 applications processed last month that don’t meet this customer
requirement. Thus, the DPMO = 150/1000 × 1,000,000, or 150,000 loans out of every million processed that fail to meet a CCR. Put differently, it means that only 850,000 loans out of a million are approved within time expectations. Statistically, 15 percent of the loans are defective and 85 percent are
correct. This is a case where all the loans processed in less than 10 days meets our criteria. Often there are
upper and lower customer requirements rather than just a single upper requirement as we have here.
•
There are two aspects to Six-Sigma programs: the methodology side and the people side.
We will take these up in order.
SIX-SIGMA METHODOLOGY
DMAIC
PCDA cycle
Continuous improvement
Kaizen
While Six Sigma’s methods include many of the statistical tools that were employed in other
quality movements, here they are employed in a systematic project-oriented fashion through
the define, measure, analyze, improve, and control (DMAIC) cycle. The DMAIC cycle is a
more detailed version of the Deming PCDA cycle, which consists of four steps—plan, do,
check, and act—that underly continuous improvement. (Continuous improvement, also
called kaizen, seeks continual improvement of machinery, materials, labor utilization, and
production methods through applications of suggestions and ideas of company teams.) Like
Six Sigma, it also emphasizes the scientific method, particularly hypothesis testing about the
relationship between process inputs (X’s) and outputs (Y’s) using design of experiments
(DOE) methods. The availability of modern statistical software has reduced the drudgery of
analyzing and displaying data and is now part of the Six-Sigma tool kit. The overarching focus
of the methodology, however, is understanding and achieving what the customer wants, since
that is seen as the key to profitability of a production process. In fact, to get across this point,
some use the DMAIC as an acronym for “Dumb Managers Always Ignore Customers.”
The standard approach to Six-Sigma projects is the DMAIC methodology developed by
General Electric, described below:3
1
2
Define (D)
• Identify customers and their priorities.
• Identify a project suitable for Six-Sigma efforts based on business objectives as
well as customer needs and feedback.
• Identify CTQs (critical-to-quality characteristics) that the customer considers to
have the most impact on quality.
Measure (M)
• Determine how to measure the process and how it is performing.
• Identify the key internal processes that influence CTQs and measure the defects
currently generated relative to those processes.
TOTAL QUALITY MANAGEMENT: FOCUS ON SIX SIGMA
3
Analyze (A)
• Determine the most likely causes of defects.
• Understand why defects are generated by identifying the key variables that are
most likely to create process variation.
4 Improve (I)
• Identify means to remove the causes of defects.
• Confirm the key variables and quantify their effects on the CTQs.
• Identify the maximum acceptance ranges of the key variables and a system for
measuring deviations of the variables.
• Modify the process to stay within an acceptable range.
5 Control (C)
• Determine how to maintain the improvements.
• Put tools in place to ensure that the key variables remain within the maximum acceptance ranges under the modified process.
A N A LY T I C A L TO O L S F O R S I X S I G M A
CONTINUOUS IMPROVEMENT
AND
The analytical tools of Six Sigma have been used for many years in traditional quality improvement programs. What makes their application to Six Sigma unique is the integration of
these tools in a corporatewide management system. The tools common to all quality efforts,
including Six Sigma, are flowcharts, run charts, Pareto charts, histograms, checksheets,
cause-and-effect diagrams, and control charts. Examples of these, along with an opportunity
flow diagram, are shown in Exhibit 7.6 arranged according to DMAIC categories where
they commonly appear.
Flowcharts. There are many types of flow charts. The one shown in Exhibit 7.6 depicts
the process steps as part of a SIPOC (supplier, input, process, output, customer) analysis.
SIPOC in essence is a formalized input-output model, used in the define stage of a project.
Run charts. They depict trends in data over time, and thereby help to understand the
magnitude of a problem at the define stage. Typically, they plot the median of a process.
Pareto charts. These charts help to break down a problem into the relative contributions of its components. They are based on the common empirical finding that a large percentage of problems are due to a small percentage of causes. In the example, 80 percent of
customer complaints are due to late deliveries, which are 20 percent of the causes listed.
Checksheets. These are basic forms that help standardize data collection. They are
used to create histograms such as shown on the Pareto chart.
Cause-and-effect diagrams. Also called fishbone diagrams, they show hypothesized
relationships between potential causes and the problem under study. Once the C&E diagram is constructed, the analysis would proceed to find out which of the potential causes
were in fact contributing to the problem.
Opportunity flow diagram. This is used to separate value-added from non-valueadded steps in a process.
Control charts. These are time-sequenced charts showing plotted values of a statistic
including a centerline average and one or more control limits. It is used here to assure
that changes introduced are in statistical control. See the technical note following this
chapter for a discussion of the various types and uses of charts for process control.
Other tools that have seen extensive use in Six-Sigma projects are failure mode and effect
analysis (FMEA) and design of experiments (DOE).
Failure mode and effect analysis. This is a structured approach to identify, estimate,
prioritize, and evaluate risk of possible failures at each stage of a process. It begins with
identifying each element, assembly, or part of the process and listing the potential failure
modes, potential causes, and effects of each failure. A risk priority number (RPN) is calculated for each failure mode. It is an index used to measure the rank importance of the
chapter 7
281
Analytical Tools for Six Sigma and Continuous Improvement
Flow Chart of Major Steps in a Process*
SUPPLIERS
INPUTS
Manufacturer
Copier
Office Supply
Company
Paper
PROCESSES
Toner
Yourself
Original
Power
Company
Electricity
OUTPUTS
CUSTOMERS
Copies
You
File
Others
Making a
Photocopy
Define
PROCESS STEPS
Put original
on glass
Close
Lid
Adjust
Settings
Press
START
Run Chart**
Average monthly volume of deliveries
(per shop)
2700
Remove
originals
and copies
DATA COLLECTION FORMS*
Checksheets are basic forms that help standardize data collection
by providing specific spaces where people should record data.
Defines what data
are being collected
Machine Downtime
(Line 13)
2400
Wendy
Operator: __________
2100
Reason
1,951 deliveries
1800
Carton Transport
1500
Metal Check
600
300
Bad Product
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
2500
75
(1890)
1500
50
1000
25
(206)
(117)
(87)
Cold food
Taste
Other
0
(220)
Wrong order
500
May want to add
space for tracking
stratification factors
Has room for
comments
100%
2000
Burned flakes
Low weight
Other
Includes place to
put the data
Pareto Chart**
Types of customer complaints
Total ⫽ 2520 October—December
(across 6 shops)
Measure
Comments
Lists the
Sealing Unit
characteristics
or conditions Barcoding
of interest
Conveyor Belt
900
0
May 19
Date: __________
Frequency
No Product
1200
Late deliveries
EXHIBIT 7.6
PRODUCT DESIGN AND PROCESS SELECTION
Unit volume
section 2
Total # of customer complaints
282
Illustration note: Delivery time was defined
by the total time from when the order was
placed to when the customer received it.
*SOURCE: RATH & STRONG, RATH & STRONG’S SIX SIGMA POCKET GUIDE, 2001.
**SOURCE: RAYTHEON SIX SIGMA, THE MEMORY JOGGER™II, 2001.
TOTAL QUALITY MANAGEMENT: FOCUS ON SIX SIGMA
C & E/Fishbone Diagram**
Reasons for late pizza deliveries
Machinery/Equipment
No capacity for
peak periods
Ovens too small
High turnover
People
Unreliable cars
Low pay
No money for repairs
Kids own junkers
No teamwork
No training
Don’t know town
People don’t show up
High
Low pay
turnover
High turnover
Drivers get lost
Rushed
Poor training Get wrong
Late pizza
information
deliveries on
Fridays &
Saturdays
Run out of ingredients
High turnover
Poor use
of space
Inaccurate
ordering
Lack of
training
Poor
training
Poor use
of space
Analyze
Poor handling of
large orders
Don’t know
town
High
turnover
Poor
Lack of experience
dispatching
Many new streets
High turnover
Methods
Materials
Opportunity Flow Diagram*
Organized to separate value-added steps from non-value-added steps.
Value-Added
Steps that are essential
even when everything
works correctly move
down the left side
Non-Value-Added
Steps that would not be needed if everything worked right the first time move horizontally
across the right side
YES
Copier YES
in Use?
Take
Original
Wait?
NO
Leave
NO
Place
Original
NO
Glass
Dirty?
YES
Clean
Improve
Select
Size
Select
Orientation
Select
Number
Paper?
NO
Box
Open?
Find
Paper
NO
YES
YES
YES
YES
Paper
Loaded?
NO
Control Chart Features*
Basic features same as a time plot
100
UCL
90
80
70
60
Control
50
40
30
20
LCL
10
0
J A S O N D J F M A M J
Control limits (calculated
from data) added to plot
Knife?
J A S O N D J F M
Centerline usually average
instead of median
Find
Help
NO
Find
Knife
Open Box
chapter 7
283
FMEA Form
FMEA Analysis
_________(revised)
Item or Potential Potential
Process Failure Effects of
Step
Mode
Failure
Potential
Cause(s)
Current
Controls
Recommended
Action
Total Risk Priority Number:
Responsibility
and
Target Date
“After”
Action Taken
Occurrence
Detection
RPN
Date: ________(original)
Team: ___________________
Severity
Project: __________________
Detection
RPN
EXHIBIT 7.7
PRODUCT DESIGN AND PROCESS SELECTION
Occurrence
section 2
Severity
284
“After” Risk Priority Number:
SOURCE: RATH & STRONG, RATH & STRONG’S SIX SIGMA POCKET GUIDE, 2001, P. 31.
items listed in the FMEA chart. See Exhibit 7.7. These conditions include the probability
that the failure takes place (occurrence), the damage resulting from the failure (severity),
and the probability of detecting the failure in-house (detection). High RPN items should
be targeted for improvement first. The FMEA suggests a recommended action to eliminate the failure condition by assigning a responsible person or department to resolve the
failure by redesigning the system, design, or process and recalculating the RPN.
Design of experiments (DOE). DOE, sometimes referred to as multivariate testing, is
a statistical methodology used for determining the cause-and-effect relationship between
process variables (X’s) and the output variable (Y). In contrast to standard statistical
tests, which require changing each individual variable to determine the most influential
one, DOE permits experimentation with many variables simultaneously through carefully selecting a subset of them.
SIX-SIGMA ROLES
AND
RESPONSIBILITIES
Successful implementation of Six Sigma is based on using sound personnel practices as
well as technical methodologies. The following is a brief summary of the personnel practices that are commonly employed in Six-Sigma implementation.
1
Executive leaders, who are truly committed to Six Sigma and who promote
it throughout the organization, and champions, who take ownership of the
processes that are to be improved. Champions are drawn from the ranks of the
TOTAL QUALITY MANAGEMENT: FOCUS ON SIX SIGMA
chapter 7
285
breakthrough
B
R E A K T H R O U G H
WHAT MAKES A GOOD CHAMPION?
At a manufacturing company implementing Six Sigma, a designated champion regularly met with his black belts. At one
report-out meeting, a black belt informed him that she needed
to purchase and install a table for sorting defects off-line. It
would cost about $17,000, but it would provide an alternative
to shutting down the entire line, which would cost far more.
The controller told her to go through the normal requisition
process and she’d have her table in about four months. That
delay would have killed the project right then and there: to
submit the project to “business as usual” would have shown
little real commitment to supporting Six Sigma. So the champion asked for the data that backed up her request, analyzed
it, agreed with it, and then got immediate executive sign-off on
securing a table the following week.
This is the stuff of a good champion: removing barriers and
sending a clear signal that he and upper management are aligned
and committed to Six Sigma. The champion does whatever it
takes to support the black belts.
SOURCE: GREG BRUE, SIX SIGMA FOR MANAGERS (NEW YORK: MCGRAW-HILL, 2002), P. 84.
EXHIBIT 7.8
B
A
Master
SPONSOR/
CHAMPION Black Belt
SPONSOR/
CHAMPION
Oversee/Guide
Project(s)
Organizing the Roles Needed
to Support Six-Sigma Efforts
Coach/Support
Project Leader
Master
Black Belt
Black Belt
Black
Belt or
Green Belt
Green Belt
or Team
Leader
Lead Project
to Success
Analyze &
Implement
Improvement
Improvement
Team
Improvement
Team
SOURCE: PETER S. PANDE, ROBERT P. NEUMAN, AND ROLAND R. CAVANAGH, THE SIX SIGMA WAY TEAM FIELDBOOK
(NEW YORK: MCGRAW-HILL, 2002), P. 31.
2
executives and managers are expected to identify appropriate metrics early in the
project and make certain that the improvement efforts focus on business results.
(See the Breakthrough box “What Makes a Good Champion?”)
Corporatewide training in Six-Sigma concepts and tools. GE spent over a billion dollars training its professional workforce in the concepts. Now, virtually every
professional in the organization is qualified in Six-Sigma techniques. To convey the
need to vigorously attack problems, professionals are given martial arts titles reflecting their skills and roles: black belts, who coach or actually lead a Six-Sigma
improvement team; master black belts, who receive in-depth training on statistical
tools and process improvement (they perform many of the same functions as black
belts but for a larger number of teams); and green belts, who are employees who
have received enough Six-Sigma training to participate in a team or, in some companies, to work individually on a small-scale project directly related to their own
job. Different companies use these “belts” in different combinations with sponsors
and champions to guide teams. Several options are shown in Exhibit 7.8.
Black belts
Master black belts
Green belts
286
section 2
PRODUCT DESIGN AND PROCESS SELECTION
3
4
Setting stretch objectives for improvement.
Continuous reinforcement and rewards. At GE, before any savings from a project
are declared, the black belt in charge must provide proof that the problems are fixed
permanently.
THE SHINGO SYSTEM: FAIL-SAFE DESIGN
● ● ● The Shingo system developed in parallel and in many ways in conflict with
the statistically based approach to quality control. As we discussed in Chapter 6 relating
to service applications, this system—or, to be more precise, philosophy of production
management—is named after the codeveloper of the Toyota just-in-time system, Shigeo
Shingo. Two aspects of the Shingo system in particular have received great attention. One
is how to accomplish drastic cuts in equipment setup times by single-minute exchange of
die (SMED) procedures. The other, the focus of this section, is the use of source inspection
and the poka-yoke system to achieve zero defects.
Shingo has argued that SQC methods do not prevent defects. Although they provide
information to tell us probabilistically when a defect will occur, they are after the fact. The
way to prevent defects from coming out at the end of a process is to introduce controls within
the process. Central to Shingo’s approach is the difference between errors and defects. Defects arise because people make errors. Even though errors are inevitable, defects can be prevented if feedback leading to corrective action takes place immediately after the errors are
made. Such feedback and action require inspection, which should be done on 100 percent of
the items produced. This inspection can be one of three types: successive check, self-check,
and source inspection. Successive check inspection is performed by the next person in the
process or by an objective evaluator such as a group leader. Information on defects is immediate feedback for the worker who produced the product, who then makes the repair. Selfcheck is done by the individual worker and is appropriate by itself on all but items that require
sensory judgment (such as existence or severity of scratches, or correct matching of shades
of paint). These require successive checks. Source inspection is also performed by the individual worker, except instead of checking for defects, the worker checks for the errors that
will cause defects. (See Exhibit 7.9 for sources of defects attributable to the worker.) This
prevents the defects from ever occurring and, hence, requiring rework. All three types of
inspection rely on controls consisting of fail-safe procedures or devices (called pokayoke). Poka-yoke includes such things as checklists or special tooling that (1) prevents the
worker from making an error that leads to a defect before starting a process or (2) gives rapid
feedback of abnormalities in the process to the worker in time to correct them.
There is a wide variety of poka-yokes, ranging from kitting parts from a bin (to ensure that
the right number of parts are used in assembly) to sophisticated detection and electronic signaling devices. An example taken from the writings of Shingo is shown in Exhibit 7.10.
There is a good deal more to say about the work of Shingo. Blasting industry’s preoccupation with control charts, Shingo states they are nothing but a mirror reflecting current conditions. When a chemical plant QC manager proudly stated that it had 200 charts in a plant of 150
people, Shingo asked him if “they had a control chart for control charts.”4 In addition to his insights into the quality area, his work on SMED is must reading for manufacturing executives.
Fail-safe procedures
Poka-yoke
ISO 9000
Glob
ISO 9000
al
www.iso.ch
ISO 14000
● ● ● ISO 9000 is a series of standards agreed upon by the International Organization for Standardization (ISO) and adopted in 1987. More than 100 countries now recognize the 9000 series for quality standards and certification for international trade. ISO 9000
evolved in Europe; in the European Common Market (ECM) alone, almost 50,000 companies have been certified as complying with these standards.
In addition to the ISO 9000 series, there is also the ISO 14000 series, which was developed to control the impact of an organization’s activities and outputs on the environment. The
ISO 14000 standards can lead to benefits such as reducing the cost of waste management, conserving energy and materials, lowering distribution costs, and improving corporate image.
TOTAL QUALITY MANAGEMENT: FOCUS ON SIX SIGMA
chapter 7
EXHIBIT 7.9
What Are the Sources of Defects?
There are various types of defects. In order of importance these are
1. Omitted processing
Sources of Defects
6. Processing wrong workpiece
2. Processing errors
7. Misoperation
3. Errors setting up workpieces
8. Adjustment error
4. Missing parts
9. Equipment not set up properly
5. Wrong parts
10. Tools and jigs improperly prepared
What are the connections between these defects and the mistakes people make?
Causal connections between
defects and human errors
Omitted processing
Processing errors
Errors setting up
workpieces
Missing parts
Wrong parts
Processing wrong
workpiece
Misoperation
Adjustment error
Improper equipment
setup
Improper tools
and jigs
SOURCE: N. K. SHIMBUN, LTD./FACTORY MAGAZINE (ED.), POKA-YOKE: IMPROVING PRODUCT QUALITY BY PREVENTING DEFECTS (CAMBRIDGE,
MA: PRODUCTIVITY PRESS, 1989), P. 14. FROM POKA-YOKE: IMPROVING PRODUCT QUALITY BY PREVENTING DEFECTS, EDITED BY NKS/FACTORY
MAGAZINE. COPYRIGHT © 1987 PRODUCTIVITY, INC, PO BOX 13390, PORTLAND, OR 97213. 800-394-6868. www.productivityinc.com
THE ISO 9000 SERIES
ISO 9000 consists of five primary parts numbered as 9000 through 9004. If we were to
display them on a continuum of an operating firm, the series would range from design and
development through procurement, production, installation, and servicing (Exhibit 7.11).
Whereas ISO 9000 and 9004 only establish guidelines for operation, ISO 9001, 9002, and
9003 are well-defined standards.
SURPRISE
Connected
NONSUPERVISION
SLOWNESS
INADVERTENT
WILLFULL
AMATEURS
MISIDENTIFICATION
FORGETFUL
CAUSES OF
DEFECTS
MISUNDERSTANDING
HUMAN
ERRORS
INTENTIONAL
Strongly connected
287
288
section 2
PRODUCT DESIGN AND PROCESS SELECTION
EXHIBIT 7.10
Before Improvement
Poka-Yoke Example (Placing
labels on parts coming
down a conveyor)
The operation depended on the
worker’s vigilance.
After Improvement
Device to ensure attachment
of labels
Labeler
Label
Blank tape
Photoelectric
tube
The tape fed out by the labeler turns sharply so that the
labels detach and project out from the tape. This is
detected by a photoelectric tube and, if the label is not
removed and applied to the product within the tact time
of 20 seconds, a buzzer sounds and the conveyor stops.
Effect: Label application failures were eliminated.
Cost: 15,000 ($75)
EXHIBIT 7.11
ISO 9000 Standards, Their
Areas of Application in
Production Flow, and
Guidelines for Use
QUALITY SYSTEM
9001: Model for Quality Assurance in Design, Production, Installation, and Servicing. (To be used when
conformance to specified requirements is to be assured by the supplier during several stages that may
include design/development, production, installation, and servicing.)
9002: Model for Quality Assurance in Procurement, Production, and Installation. (To be used when conformance
to specified requirements is to be assured by the supplier during production and installation.)
9003: Model for Quality Assurance in Final Inspection Test. (To be used when conformance to specified
requirements is to be assured by the supplier solely at final inspection and test.)
GUIDELINES FOR USE
9000: Quality Management and Quality Assurance Standards—Guidelines for Selection and Use.
9004: Quality Management and Quality System Elements—Guidelines.
Design/
Development
Procurement
Production
Installation
Servicing
ISO9003
ISO9002
ISO9001
Quite a bit of work and expense may be needed to be accredited at the highest level,
which is 9001. Furthermore, some firms may not need ISO 9001 accreditation. For example,
note that in Exhibit 7.11, ISO 9003 covers quality in production’s final inspection and testing. A firm can be accredited at this level of final production only. This would essentially
guarantee the firm’s quality of final output and be attractive to customers. A broader accreditation would be 9002, which extends from purchasing and production through installation.
TOTAL QUALITY MANAGEMENT: FOCUS ON SIX SIGMA
chapter 7
There are 20 elements in the ISO 9000 standards that relate to how the system operates
and how well it is performing. These are contained in section 4 of the ISO 9000 Guidelines.
Each of these elements applies in varying degrees to the three standards 9001, 9002, and
9003. (ISO 9001 contains all of them.)
ISO 9000 is intentionally vague. (For example, a sample standard statement is “Procedures
shall be prepared.”) A firm interprets the requirements as they relate to its business. From a
practical and useful standpoint for businesses, ISO 9000 is valuable to firms because it provides a framework so they can assess where they are and where they would like to be. In its
simplest terms, it is sometimes stated that ISO 9000 directs you to “document what you do
and then do as you documented.”
I S O 9 0 0 0 C E RT I F I C AT I O N
Why is it important to become ISO 9000 certified? For one reason, it is essential from a
purely competitive standpoint. Consider the situation where you need to purchase parts for
your firm, and several suppliers offer similar parts at similar prices. Assume that one of
these firms has been ISO 9000 certified and the others have not. From whom would you
purchase? There is no doubt that the ISO 9000–certified company would have the inside
track in your decision making. Why? Because ISO 9000 specifies the way the supplier firm
operates as well as its quality standards, delivery times, service levels, and so on.
There are three forms of certification:
1 First party: A firm audits itself against ISO 9000 standards.
2 Second party: A customer audits its supplier.
3 Third party: A “qualified” national or international standards or certifying agency
serves as auditor.
The best certification of a firm is through a third party. Once passed by the third-party audit,
a firm is certified and may be registered and recorded as having achieved ISO 9000 status,
and it becomes part of a registry of certified companies. This third-party certification also
has legal advantages in the European Community. For example, a manufacturer is liable
for injury to a user of the product. The firm, however, can free itself from any liability by
showing that it has used the appropriate standards in its production process and carefully
selected its suppliers as part of its purchasing requirement. For this reason, there is strong
motivation to choose ISO 9000–certified suppliers.
EXTERNAL BENCHMARKING FOR
QUALITY IMPROVEMENT
● ● ● The quality improvement approaches described so far are more or less inward
looking. They seek to make improvements by analyzing in detail the current practices of
the company itself. External benchmarking, however, goes
outside the organization to examine what industry competitors
and excellent performers outside of industry are doing. Benchmarking typically involves the following steps:
Identify processes needing improvement. Identify a firm
that is the world leader in performing the process. For many
processes, this may be a company that is not in the same
industry. Examples would be Proctor & Gamble using
L.L Bean as the benchmark in evaluating its order entry system, or ICL (a large British computer maker) benchmarking Marks and Spenser (a large U.K. clothing retailer) to
improve its distribution system. A McKinsey study cited a
firm that measured pit stops on a motor racing circuit as
a benchmark for worker changes on its assembly line.5
External benchmarking
289
290
section 2
PRODUCT DESIGN AND PROCESS SELECTION
Contact the managers of that company and make a personal visit to interview managers
and workers. Many companies select a team of workers from that process as part of the
team of visitors.
Analyze data. This entails looking at gaps between what your company is doing and
what the benchmarking company is doing. There are two aspects of the study: one is comparing the actual processes; the other is comparing the performance of these processes
according to a set of measures. The processes are often described using flowcharts and
subjective evaluations of how workers relate to the process. In some cases, companies
permit videotaping, although there is a tendency now for benchmarked companies to keep
things under wraps for fear of giving away process secrets.
SERV
SERVICE QUALITY MEASUREMENT:
SERVQUAL
AL
QU
SERVQUAL
● ● ● Although the approaches to improving product quality are equally applicable
to services, identifying what should be improved requires tapping the customer’s satisfaction
with the service process as well as the outcome from that process. A standard approach to
making this determination is to measure the gap between what customers expected and their
perceptions of the service provided in a service encounter (see Exhibit 7.12). The size of the
gap indicates where improvements should be made. The measurement is done by having
customers fill out the SERVQUAL questionnaire6 (contained on the CD-ROM that accompanies this book). The questionnaire consists of 22 expectation and matching perception
questions relating to the five statistically derived dimensions of service quality listed in
Exhibit 7.12. Each item is scored on a 1 to 7 scale, from strongly disagree (1) to strongly agree
(7). For example, expectation question 1 states that the company under study “should have
up-to-date equipment”; and perception question 1 states that the company “has up-to-date
equipment.” Thus, if the customer assigns a 6 to expectation and 4 to perception, the gap
score is −2. (By convention, expectations are subtracted from perceptions.) Typically, tangibles have the lowest gap score because physical features of a service are easier to control.
W e i g h t e d S E R V Q U A L The SERVQUAL procedure also permits assignment of
importance weightings to each of the five dimensions. This is done by allocating 100 points
across the dimensions and then multiplying the dimension gap score by points assigned to it.
EXHIBIT 7.12
Perceived Service Quality
Word of
mouth
Dimensions of
Service Quality*
Reliability
Responsiveness
Assurance
Empathy
Tangibles
Personal
needs
Expected
service
Perceived
service
Past
experience
Perceived Service Quality
1. Expectations exceeded
ES < PS (Quality surprise)
2. Expectations met
ES ≈ PS (Satisfactory quality)
3. Expectations not met
ES > PS (Unacceptable quality)
*Reliability: the ability to perform service as promised, both dependably and accurately.
Responsiveness: willingness to help customers promptly.
Assurance: knowledge and courtesy of employees, as well as their ability to convey trust.
Empathy: caring and individualized attention.
Tangibles: the appearance of physical facilities, equipment, and personnel, as well as other factors affecting the senses such as
noise and temperature.
SOURCE: ADAPTED WITH PERMISSION FROM THE JOURNAL OF MARKETING, PUBLISHED BY THE AMERICAN MARKETING ASSOCIATION, A. PARASURAMAN,
V. A. ZEITHHAML, AND L. L. BERRY, “A CONCEPTUAL MODEL OF SERVICE QUALITY AND ITS IMPLICATIONS FOR FUTURE RESEARCH,” FALL 1985/49, P. 48.
TOTAL QUALITY MANAGEMENT: FOCUS ON SIX SIGMA
1. Reliability involves the correct technical functioning of the site and the accuracy of service promises
(having items in stock, delivering when promised), billing, and product information.
2. Responsiveness means quick response and the ability to get help if there is a problem or question.
chapter 7
291
EXHIBIT 7.13
Dimensions of Perceived e-SQ
3. Access is the ability to get on the site quickly and to reach the company when needed.
4. Flexibility involves choice of ways to pay, ship, buy, search for, and return items.
5. Ease of navigation means that a site contains functions that help customers find what they need without
difficulty, possesses a good search engine, and allows the customer to maneuver easily and quickly back
and forth through the pages.
6. Efficiency means that a site is simple to use, is structured properly, and requires a minimum of information
to be input by the customer.
7. Assurance trust involves the confidence the customer feels in dealing with the site and is due to the reputation of the site and the products or services it sells as well as clear and truthful information presented.
8. Security privacy involves the degree to which the customer believes the site is safe from intrusion and
personal information is protected.
9. Price knowledge is the extent to which the customer can determine shipping price, total price, and
comparative prices during the shopping process.
10. Site aesthetics relates to the appearance of the site.
11. Customization/personalization is how much and how easily the site can be tailored to individual
customers’ preferences, histories, and ways of shopping.
SOURCE: VALARIE A. ZEITHAML, A. PARASURAMAN, AND ARVIND MALHOTRA, “A CONCEPTUAL FRAMEWORK FOR UNDERSTANDING E-SERVICE QUALITY:
IMPLICATIONS FOR FUTURE RESEARCH AND MANAGERIAL PRACTICE,” MARKETING SCIENCE INSTITUTE, REPORT SUMMARY 00-115, 2001, P. 14.
All five weighted dimension scores are then added to derive an overall weighted service quality score. Bank customers, for example, usually weight reliability heavily and tangibles lightly.
e - S E R V I C E Q U A L I T Y A new version of SERVQUAL, e-SERVICE QUALITY
has been developed to evaluate service on the Internet. e-SQ is defined as the extent to
which a website facilitates efficient and effective shopping, purchasing, and delivery. Its
dimensions are shown in Exhibit 7.13.
e-SERVICE QUALITY
CONCLUSION
● ● ● How to achieve TQM is no secret anymore. The challenge is to make certain
that a quality program really does have a customer focus and is sufficiently agile to be able
to make improvements quickly without losing sight of the real time needs of the business.
The quality system must be analyzed for its own quality. There is also a need for sustaining
a quality culture over the long haul. Some companies (who will remain nameless) that
gained a great reputation for quality in the 1980s and ’90s simply ran out of gas in their
quality efforts—their managers just couldn’t sustain the level of enthusiasm necessary for
quality to remain a top priority goal. As Tom Peters has said, “Most Quality programs fail
for one of two reasons: they have system without passion, or passion without system.”7
KEY TERMS
Total quality management (TQM) Managing the entire organization
so that it excels on all dimensions of products and services that are
important to the customer.
Malcolm Baldrige National Quality Award An award established by
the U.S. Department of Commerce given annually to companies that
excel in quality.
Design quality The inherent value of the product in the marketplace.
Conformance quality The degree to which the product or service
design specifications are met.
Quality at the source The person who does the work is responsible
for ensuring that specifications are met.
Dimensions of quality Criteria by which quality is measured.
Cost of quality Expenditures related to achieving product or service
quality, such as the costs of prevention, appraisal, internal failure, and
external failure.
Six Sigma A statistical term to describe the quality goal of no more
than four defects out of every million units. Also refers to a quality
improvement philosophy and program.
292
section 2
PRODUCT DESIGN AND PROCESS SELECTION
DPMO (defects per million opportunities) A metric used to describe
the variability of a process.
Fail-safe or poka-yoke procedures Simple practices that prevent
errors or provide feedback in time for the worker to correct errors.
DMAIC An acronym for the Define, Measure, Analyze, Improve, and
Control improvement methodology followed by companies engaging in Six-Sigma programs.
ISO 9000 Formal standards used for quality certification, developed
by the International Organization for Standardization.
PDCA cycle Also called “The Deming cycle or wheel”; refers to the
plan–do–check–act cycle of continuous improvement.
Continuous improvement The philosophy of continually seeking
improvements in processes through the use of team efforts.
Kaizen Japanese term for continuous improvement.
Black belts, master black belts, green belts Terms used to describe
different levels of personal skills and responsibilities in Six-Sigma
programs.
REVIEW
AND
ISO 14000 A series of standards to control the impact of an organization’s activities and outputs on the environment.
External benchmarking Looking outside the company to examine
what excellent performers inside and outside the company’s industry are doing in the way of quality.
SERVQUAL A service quality questionnaire that measures the gap
between customer expectations and perceptions of performance after
a service encounter.
e-SERVICE QUALITY A version of SERVQUAL designed to evaluate
service on the Internet.
DISCUSSION QUESTIONS
1 Is the goal of Six Sigma realistic for services such as Blockbuster Video stores?
2 “If line employees are required to work on quality improvement activities, their productivity
will suffer.” Discuss.
3 “You don’t inspect quality into a product; you have to build it in.” Discuss the implications of
this statement.
4 “Before you build quality in, you must think it in.” How do the implications of this statement
differ from those in question 4?
5 Business writer Tom Peters has suggested that in making process changes, we should “Try it,
test it, and get on with it.” How does this square with the DMAIC/continuous improvement
philosophy?
6 Shingo told a story of a poka-yoke he developed to make sure that the operators avoided the
mistake of putting fewer than the required four springs in a push-button device. The existing
method involved assemblers taking individual springs from a box containing several hundred,
and then placing two of them behind an ON button and two more behind an OFF button. What
was the poka-yoke Shingo created?
7 A typical word processing package is loaded with poka-yokes. List three. Are there any others
you wish the packages had?
PROBLEMS
1 A manager states that his process is really working well. Out of 1,500 parts, 1,477 were produced free of a particular defect and passed inspection. Based upon DPMO, how would you
rate this performance, other things being equal?
2 Professor Chase is frustrated by his inability to make a good cup of coffee in the morning.
Show how you would use a fishbone diagram to analyze the process he uses to make a cup of
his evil brew.
3 Use the benchmarking process and as many DMAIC/CI analytical tools as you can to show
how you can improve your performance in your weakest course in school.
4 Consider a simple repair job that you performed that did not turn out particularly well. Analyze
the mistakes or defects you made using the Sources of Defects table presented in Exhibit 7.9.
Which errors were your fault?
5 Evaluate the service quality of your bank using the perceptions of performance portion of the
SERVQUALquestionnaire contained on the student CD-ROM. Compute the score for each of the
five SERVQUAL dimensions. On which dimensions does your bank score the highest? Lowest?
6 Evaluate the service quality of your bank’s website using the Dimensions of Perceived e-SQ
given in Exhibit 7.13.
7 Prepare a SIPOC flowchart of the major steps in the process of boarding a commercial flight.
Start the process with the passenger arriving curbside at your local airport.
8 Prepare an opportunity flow diagram for the same process of boarding a commercial flight.
TOTAL QUALITY MANAGEMENT: FOCUS ON SIX SIGMA
1
2
XERCISES
Visit the Baldrige Award website and see who won this year. What quality ideas did the winner
demonstrate? What did the winner do that was particularly creative?
Visit the Six Sigma website to see how companies are applying the concept.
CASE:
HANK KOLB, DIRECTOR
OF
293
www.quality.nist.gov/show.htm
net
Inte
r
INTERNET ENRICHMENT E
chapter 7
www.isixsigma.com
QUALITY ASSURANCE
● ● ● Hank Kolb was whistling as he walked toward his
office, still feeling a bit like a stranger since he had been hired four
weeks before as director of quality assurance. All that week he had
been away from the plant at an interesting seminar, entitled “Quality
in the 90s,” given for quality managers of manufacturing plants by
the corporate training department. He was now looking forward to
digging into the quality problems at this industrial products plant
employing 1,200 people.
Kolb poked his head into the office of Mark Hamler, his immediate subordinate as the quality control manager, and asked him how
things had gone during the past week. Hamler’s muted smile and an
“Oh, fine,” stopped Kolb in his tracks. He didn’t know Hamler very
well and was unsure about pursuing this reply any further. Kolb was
still uncertain of how to start building a relationship with him because Hamler had been passed over for the promotion to Kolb’s job;
Hamler’s evaluation form had stated “superb technical knowledge;
managerial skills lacking.” Kolb decided to inquire a little further
and asked Hamler what had happened; he replied, “Oh, just another
typical quality snafu. We had a little problem on the Greasex line
last week [a specialized degreasing solvent packed in a spray can
for the high-technology sector]. A little high pressure was found in
some cans on the second shift, but a supervisor vented them so that
we could ship them out. We met our delivery schedule!” Because
Kolb was still relatively unfamiliar with the plant and its products,
he asked Hamler to elaborate; painfully, Hamler continued:
We’ve been having some trouble with the new filling
equipment and some of the cans were pressurized beyond
our AQL [acceptable quality level] on a psi rating scale.
The production rate is still 50 percent of standard, about
14 cases per shift, and we caught it halfway into the shift.
Mac Evans [the inspector for that line] picked it up, tagged
the cases “hold,” and went on about his duties. When he returned at the end of the shift to write up the rejects, Wayne
Simmons, first-line supervisor, was by a pallet of finished
goods finishing sealing up a carton of the rejected Greasex;
the reject “hold” tags had been removed. He told Mac that
he had heard about the high pressure from another inspector at coffee break, had come back, taken off the tags, individually turned the cans upside down and vented every one
of them in the eight rejected cartons. He told Mac that production planning was really pushing for the stuff and they
couldn’t delay by having it sent through the rework area.
He told Mac that he would get on the operator to run the
equipment right next time. Mac didn’t write it up but came
in about three days ago to tell me about it. Oh, it happens
every once in a while and I told him to make sure to check
with maintenance to make sure the filling machine was
adjusted; and I saw Wayne in the hall and told him that he
ought to send the stuff through rework next time.
Kolb was a bit dumbfounded at this and didn’t say much—he
didn’t know if this was a big deal or not. When he got to his office
he thought again what Morganthal, general manager, had said when
he had hired him. He warned Kolb about the “lack of quality attitude”
in the plant, and said that Kolb “should try and do something about
this.” Morganthal further emphasized the quality problems in the
plant: “We have to improve our quality; it’s costing us a lot of money,
I’m sure of it, but I can’t prove it! Hank, you have my full support in
this matter; you’re in charge of these quality problems. This downward quality–productivity–turnover spiral has to end!”
The incident had happened a week before; the goods were probably out in the customers’ hands by now, and everyone had forgotten
about it (or wanted to). There seemed to be more pressing problems
than this for Kolb to spend his time on, but this continued to nag
him. He felt that the quality department was being treated as a joke,
and he also felt that this was a personal slap from manufacturing. He
didn’t want to start a war with the production people, but what could
he do? Kolb was troubled enough to cancel his appointments and
spend the morning talking to a few people. After a long and very
tactful morning, he learned the following information:
1
From personnel. The operator for the filling equipment
had just been transferred from shipping two weeks ago. He
had no formal training in this job but was being trained by
Wayne, on the job, to run the equipment. When Mac had
tested the high-pressure cans, the operator was nowhere to be
found and had only learned of the rejected material from
Wayne after the shift was over.
2 From plant maintenance. This particular piece of automated filling equipment had been purchased two years ago
for use on another product. It had been switched to the Greasex line six months ago and maintenance completed 12 work
orders during the last month for repairs or adjustments on it.
The equipment had been adapted by plant maintenance for
handling the lower viscosity of Greasex, which it had not
originally been designed for. This included designing a special filling head. There was no scheduled preventive maintenance for this equipment, and the parts for the sensitive
filling head, replaced three times in the last six months, had
to be made at a nearby machine shop. Nonstandard downtime was 15 percent of actual running time.
3 From purchasing. The plastic nozzle heads for the Greasex
can, designed by a vendor for this new product on a rush order,
were often found to have slight burrs on the inside rim, and this
caused some trouble in fitting the top to the can. An increase in
application pressure at the filling head by maintenance adjustment had solved the burr application problem or had at least
forced the nozzle heads on despite burrs. Purchasing agents
said that they were going to talk to the sales representative of
the nozzle head supplier about this the next time he came in.
4 From product design and packaging. The can, designed
especially for Greasex, had been contoured to allow better
gripping by the user. This change, instigated by marketing research, set Greasex apart from the appearance of its
competitors and was seen as significant by the designers.
There had been no test of the effects of the contoured can on
294
section 2
PRODUCT DESIGN AND PROCESS SELECTION
What bothered Kolb most was the safety issue of the high pressure in the cans. He had no way of knowing how much of a hazard
the high pressure was or if Simmons had vented them enough to
effectively reduce the hazard. The data from the can manufacturer,
which Hamler had showed him, indicated that the high pressure found
by the inspector was not in the danger area. But, again, the inspector
had used only a sample testing procedure to reject the eight cases.
Even if he could morally accept that there was no product safety hazard, could Kolb make sure that this would never happen again?
Skipping lunch, Kolb sat in his office and thought about the
morning’s events. The past week’s seminar had talked about the role
of quality, productivity and quality, creating a new attitude, and the
quality challenge; but where had they told him what to do when this
happened? He had left a very good job to come here because he
thought the company was serious about the importance of quality,
and he wanted a challenge. Kolb had demanded and received a
salary equal to the manufacturing, marketing, and R&D directors,
and he was one of the direct reports to the general manager. Yet he
still didn’t know exactly what he should or shouldn’t do, or even
what he could or couldn’t do under these circumstances.
filling speed or filling hydrodynamics from a high-pressured
filling head. Kolb had a hunch that the new design was acting
as a venturi (carrier creating suction) when being filled, but
the packaging designer thought that was unlikely.
5 From the manufacturing manager. He had heard about
the problem; in fact, Simmons had made a joke about it,
bragging about how he beat his production quota to the other
foremen and shift supervisors. The manufacturing manager
thought Simmons was one of the “best foremen we have . . .
he always got his production out.” His promotion papers
were actually on the manufacturing manager’s desk when
Kolb dropped by. Simmons was being strongly considered
for promotion to shift supervisor. The manufacturing manager, under pressure from Morganthal for cost improvements
and reduced delivery times, sympathetized with Kolb but
said that the rework area would have vented with their pressure gauges what Wayne had done by hand. “But I’ll speak
with Wayne about the incident,” he said.
6 From marketing. The introduction of Greasex had been
rushed to market to beat competitors, and a major promotional
advertising campaign was under way to increase consumer
awareness. A deluge of orders was swamping the order-taking
department and putting Greasex high on the back-order list.
Production had to turn the stuff out; even being a little off spec
was tolerable because “it would be better to have it on the
shelf than not there at all. Who cares if the label is a little
crooked or the stuff comes out with a little too much pressure?
We need market share now in that high-tech segment.”
QUESTIONS
1 What are the causes of the quality problems on the Greasex
line? Display your answer on a fishbone diagram.
2 What general steps should Hank follow in setting up a continuous improvement program for the company? What problems will he have to overcome to make it work?
SOURCE: COPYRIGHT 1981 BY PRESIDENT AND FELLOWS OF HARVARD COLLEGE, HARVARD BUSINESS SCHOOL. CASE 681.083. THIS CASE WAS PREPARED BY FRANK S. LEONARD AS THE BASIS FOR CLASS DISCUSSION
REPRINTED BY PERMISSION OF THE HARVARD BUSINESS SCHOOL.
RATHER THAN TO ILLUSTRATE EITHER EFFECTIVE OR INEFFECTIVE HANDLING OF AN ADMINISTRATIVE SITUATION.
CASE:
SHORTENING CUSTOMERS’ TELEPHONE WAITING TIME
● ● ● This case illustrates how a bank applied some of the
basic tools shown in Exhibit 7.6 and storyboard concepts to improve
customer service. It is the story of a quality circle (QC) program
implemented in the main office of a large bank. An average of 500
customers call this office every day. Surveys indicated that callers
tended to become irritated if the phone rang more than five times
before it was answered, and often would not call the company again.
In contrast, a prompt answer after just two rings reassured the customers and made them feel more comfortable doing business by
phone.
SELECTION OF A THEME
Telephone reception was chosen as a QC theme for the following
reasons: (1) Telephone reception is the first impression a customer
receives from the company; (2) this theme coincided with the company’s telephone reception slogan, “Don’t make customers wait, and
avoid needless switching from extension to extension”; and (3) it
also coincided with a companywide campaign being promoted at
that time that advocated being friendly to everyone one met.
First, the staff discussed why the present method of answering
calls made callers wait. Exhibit 7.14 illustrates a frequent situation
EXHIBIT 7.14
Why Customers Had to Wait
(1)
Customer A
Customer B
(2)
Operator
Receiving
party
TOTAL QUALITY MANAGEMENT: FOCUS ON SIX SIGMA
295
chapter 7
EXHIBIT 7.15
Cause-and-Effect Diagram
Receiving party not present
Working system of operators
Absent
Telephone call rush
Out of office
Lunchtime rest
Not at desk
Absent
Makes
customer
wait
Lengthy
conversation
Not giving section
and name of
receiving party
Complaining
Does not understand
customer’s message
Lack of
knowledge of
company jobs
Takes time to explain
branch office location
Starts leaving a message
Customer
Operator
EXHIBIT 7.16
Causes of Callers’ Waits
A. Checksheet—Designed to Identify the Problems
REASON
Date
NO ONE PRESENT
IN SECTION
RECEIVING THE CALL
RECEIVING PARTY
NOT PRESENT
ONLY ONE OPERATOR
(PARTNER OUT OF
THE OFFICE)
TOTAL
June 4
24
June 5
32
June 6
28
June 15
25
B. Reasons Why Callers Had to Wait
DAILY AVERAGE
TOTAL NUMBER
14.3
172
A
One operator (partner out of the office)
B
Receiving party not present
6.1
73
C
No one present in the section receiving the call
5.1
61
D
Section and name of receiving party not given
1.6
19
E
Inquiry about branch office locations
1.3
16
F
Other reasons
0.8
10
29.2
351
Total
C. Reasons Why Callers
Had to Wait (Pareto
Diagram)
100%
87.1%
300
71.2%
200
49.0%
100
Period: 12 days from June 4 to 16, 1980
0
ABCDEF
296
section 2
PRODUCT DESIGN AND PROCESS SELECTION
in which a call from customer B comes in while the operator is talking with customer A. Let’s see why the customer has to wait.
At (1), the operator receives a call from the customer but, due to
lack of experience, does not know where to connect the call. At (2),
the receiving party cannot answer the phone quickly, perhaps because
he or she is unavailable, and no one else can take the call. The result
is that the operator must transfer the call to another extension while
apologizing for the delay.
CAUSE-AND-EFFECT (FISHBONE) DIAGRAM AND
SITUATION ANALYSIS
To fully understand the situation, the quality circle members decided
to conduct a survey regarding callers who waited for more than
five rings. Circle members itemized factors at a brainstorming discussion and arranged them in a cause-and-effect diagram. (See Exhibit 7.15.) Operators then kept checksheets on several points to tally
the results spanning 12 days from June 4 to 16. (See Exhibit 7.16.)
RESULTS OF THE CHECKSHEET SITUATION ANALYSIS
The data recorded on the checksheets unexpectedly revealed that
“one operator (partner out of the office)” topped the list by a big
margin, occurring a total of 172 times. In this case, the operator on
duty had to deal with large numbers of calls when the phones were
EXHIBIT 7.17
busy. Customers who had to wait a long time averaged 29.2 daily,
which accounted for 6 percent of the calls received every day. (See
Exhibits 7.16B and 7.16C.)
SETTING THE TARGET
After an intense but productive discussion, the staff decided to set a
QC program goal of reducing the number of waiting callers to zero.
That is to say that all incoming calls would be handled promptly,
without inconveniencing the customer.
MEASURES AND EXECUTION
1 Taking lunches on three different shifts, leaving at least
two operators on the job at all times. Until this resolution
was made, a two-shift lunch system had been employed, leaving only one operator on the job while the other was taking a
lunch break. However, because the survey revealed that this
was a major cause of customers waiting on the line, the company brought in a helper operator from the clerical section.
2 Asking all employees to leave messages when leaving their
desks. The objective of this rule was to simplify the operator’s chores when the receiving party was not at his desk.
The new program was explained at the employees’ regular
morning meetings, and companywide support was requested.
Effects of QC
A. Effects of QC (Comparison of before and after QC)
REASON WHY CALLERS HAD TO WAIT
A
B
C
D
E
F
TOTAL NUMBER DAILY AVERAGE
BEFORE AFTER BEFORE AFTER
One operator (partner out of the office)
Receiving party not present
No one present in the section receiving the call
Section and name of receiving party not given
Inquiry about branch office locations
Others
Total
172
73
61
19
16
10
15
17
20
4
3
0
14.5
6.1
5.1
1.6
1.3
0.8
1.2
1.4
1.7
0.3
0.2
0
351
59
29.2
4.8
Period: 12 days from Aug. 17 to 30.
Problems are classified according to cause and presented in order of the amount of time
consumed. They are illustrated in a bar graph. 100% indicates the total number of time-consuming calls.
B. Effects of QC (Pareto Diagram)
100%
350
300
Effects
300
200
200
Before
100
After
100
100%
0
0
A B C D E F
C B A D E
TOTAL QUALITY MANAGEMENT: FOCUS ON SIX SIGMA
3
To help implement this practice, posters were placed around
the office to publicize the new measures.
Compiling a directory listing the personnel and their
respective jobs. The notebook was specially designed to
aid the operators, who could not be expected to know the
details of every employee’s job. The notebook would help
properly route calls.
chapter 7
297
CONFIRMING THE RESULTS
Although the waiting calls could not be reduced to zero, all
items presented showed a marked improvement as shown in
Exhibits 7.17A and 7.17B. The major cause of delays, “one operator (partner out of the office),” plummeted from 172 incidents
during the control period to 15 in the follow-up survey.
SOURCE: FROM “THE QUEST FOR HIGHER QUALITY—THE DEMING PRIZE AND QUALITY CONTROL,” RICOH COMPANY, LTD., IN MASAAKI IMAI, KAIZEN: THE KEY TO JAPAN’S COMPETITIVE SUCCESS (NEW YORK:
RANDOM HOUSE, 1986), PP. 54–58.
CASE:
“ H E Y, I S A N Y B O D Y T H E R E ? ” A N E X A M P L E
AMERICAN EXPRESS
● ● ● In this case the customer is using Six Sigma to reduce
defects in a service.
THE GENERAL SITUATION
A number of merchants that accept American Express cards fail to
place point-of-purchase materials (e.g., decals) that notify customers
that they can use these cards at these establishments while displaying
competing (e.g., Visa, Mastercard, etc.) point-of-purchase materials.
American Express defines these merchants as “passive suppressors.”
In an effort to increase visibility, an external vendor that placed
point-of-purchase material in the marketplace, identified passive
suppressors, and measured placement and passive suppression rates
was hired by American Express. However, the vendor had a significant rate of failure to contact or meet with the merchants. The leading reason for not meeting with the merchant was that the store was
closed when the vendor stopped by.
OF
DMAIC
AT
uncallable without first checking to see if any point-of-purchase materials were visible from the outside. This resulted in merchants who
displayed point-of-purchase material being visited multiple times—
leading to rework.
IMPROVE
American Express then tested and validated their hypotheses. The
call hours for all visits were changed to begin after 10:00 A.M. The
vendor was required to continue the inspection process with respect
to external placement of point-of-purchase material. The first change,
revised call hours, resulted in a decrease to 4.5 percent from 8.0 percent in the defect rate. The second change, continued inspection,
indicated that 35.4 percent of the remaining 4.5 percent closed stores
actually had external point-of-purchase material displayed. Combined, these two changes had the following effects: the defect rate
decreased to 2.8 percent, the number of defects per million decreased
to 28,000, and the sigma level increased to 3.2.
DEFINE AND MEASURE
The objective was to reduce closed store uncallables (failures to
contact), which represented 27.4 percent of total uncallables and 8.0
percent of the annualized attempted visits. The process represented
a 2.9 sigma level and 80,000 defects per million opportunities.
CONTROL
In order to achieve control, American Express uses a p control chart
to track the proportion of closed stores over time and the vendor call
report was revised to reflect the uncallable rate by reason.
ANALYZE
A Pareto chart pointed to the “closed store” category as the number
one reason for uncallables. By shadowing the vendor on merchant
visits, American Express learned that the visits took place between
8:00 A.M. and 6:00 P.M. Of the closed stores, 45 percent were retail
establishments and 16 percent were restaurants. Typically, these two
types of establishments are not open before 10:00 A.M. Therefore,
American Express hypothesized that the hours the vendor was calling on the merchant contributed to the high uncallable rate. It also
was determined that if an establishment was closed, the inspection process was terminated with the merchant being reported as
QUESTIONS
1 In terms of “design quality” and “conformance quality,” explain how using the Six-Sigma approach helped American
Express.
2 In the case, American Express uncovered the two primary
causes of the uncallable rate by “shadowing” the vendor.
What Six Sigma/continuous improvement tools might the
vendor have used to uncover the same information and revise
the process?
3 What are some of the limitations of the Six-Sigma approach
when there is subjectivity in the metrics used?
SOURCE: SAI KIM, “SERVICE QUALITY SIX SIGMA CASE STUDIES,” ANNUAL QUALITY CONGRESS PROCEEDINGS 54 (MAY 2000).
SELECTED BIBLIOGRAPHY
Bemowski, K., and B. Stratton, eds. 101 Good Ideas: How to Improve
Just About Any Process. Washington, DC: American Society for
Quality, 1999.
Chowdhury, Subir. Design for Six Sigma. Chicago: Dearborn Trade Publishing, 2002.
Blakeslee, Jerome A., Jr. “Implementing the Six Sigma Solution.” Quality
Progress, July 1999, pp. 77–85.
Chowdhury, S., and K. Zimmer. QS-9000 Pioneers—Registered Companies Share Their Strategies for Success. Burr Ridge, IL: Richard D.
Irwin, 1996.
Brue, Greg. Six Sigma for Managers. New York: McGraw-Hill, 2002.
Crosby, P. B. Quality Is Free. New York: McGraw-Hill, 1979.
298
section 2
PRODUCT DESIGN AND PROCESS SELECTION
______. Quality Is Still Free. New York: McGraw-Hill, 1996.
Deming, W. E. Quality, Productivity and Competitive Position. Cambridge,
MA: MIT Center for Advanced Engineer Study, 1982.
Eckes, George. Six Sigma Revolution: How General Electric and Others
Turned Process into Profits. New York: John Wiley & Sons, 2001.
Evans, James R., and William M. Lindsay. The Management and Control of
Quality. Cincinnati: South-Western/Thomson Learning, 2002.
Pande, Peter S.; Robert P. Neuman; and Roland R. Cavanagh. The Six
Sigma Way. New York: McGraw-Hill, 2000.
______. The Six Sigma Way Team Fieldbook. New York: McGraw-Hill,
2002.
Robinson, A. Modern Approaches to Manufacturing Improvement: The
Shingo System. Cambridge, MA: Productivity Press, 1990.
Shingo, S. Zero Quality Control: Source Inspection and the Poka-Yoke
System. Stanford, CT: Productivity Press, 1986.
Feigenbaum, A. V. Total Quality Control. New York: McGraw-Hill, 1991.
Gitlow, H.; A. Oppenheim; and R. Oppenheim. Quality Management:
Tools and Methods for Improvement. 2nd ed. New York: Irwin/
McGraw-Hill, 1995.
Juran, J. M. Quality Control Handbook. 3rd ed. New York: McGraw-Hill,
1979.
Taormina, T. Virtual Leadership and the ISO 9000 Imperative. Englewood
Cliffs, NJ: Prentice Hall, 1996.
Welch, Jack. Jack: Straight from the Gut. New York: Warner Business
Books, 2001.
Zemke, R., and J. A. Woods, eds. Best Practices in Customer Service.
Amherst, NJ: HRD Press, 1999.
Juran, J. M., and F. M. Gryna. Quality Planning and Analysis. 2nd ed.
New York: McGraw-Hill, 1980.
FOOTNOTES
1
2
3
4
D. A. Garvin, Managing Quality. Free Press, 1988.
P. B. Crosby, Quality Is Free (New York: New American Library, 1979), p. 15.
S. Walleck, D. O’Halloran, and C. Leader, “Benchmarking World-Class Performance,” McKinsey Quarterly, no. 1 (1991), p. 7.
A. Robinson, Modern Approaches to Manufacturing Improvement: The Shingo System (Cambridge, MA: Productivity Press,
1990, p. 234.
5 Walleck, O’Halloran, and Leader, “Benchmarking World-Class Performance,” p. 7.
6 A. Parasuraman, V. A. Zeithaml, and L. L. Berry, “SERVQUAL: A Multiple-Item Scale for Measuring Customer Perceptions
of Service Quality,” Journal of Retailing 64, no. 1 (Spring 1988), pp. 12–40.
7 Tom Peters, Thriving on Chaos (New York: Knopf, 1987), p. 74.
`